Macroeconomic forecasting with mixed data sampling frequencies: Evidence from a small open economy
Author | Albert K. Tsui,Cheng Yang Xu,Zhaoyong Zhang |
DOI | http://doi.org/10.1002/for.2528 |
Published date | 01 September 2018 |
Date | 01 September 2018 |
RESEARCH ARTICLE
Macroeconomic forecasting with mixed data sampling
frequencies: Evidence from a small open economy
Albert K. Tsui
1
| Cheng Yang Xu
1
| Zhaoyong Zhang
2
1
Department of Economics, National
University of Singapore, Singapore
2
School of Business and Law, Edith
Cowan University, Joondalup, WA,
Australia
Correspondence
Zhaoyong Zhang, School of Business and
Law, Edith Cowan University, 270
Joondalup Drive, Joondalup, WA 6027,
Australia.
Email: zhaoyong.zhang@ecu.edu.au
Abstract
The aim of this study was to forecast the Singapore gross domestic product
(GDP) growth rate by employing the mixed‐data sampling (MIDAS) approach
using mixed and high‐frequency financial market data from Singapore, and to
examine whether the high‐frequency financial variables could better predict
the macroeconomic variables. We adopt different time‐aggregating methods
to handle the high‐frequency data in order to match the sampling rate of
lower‐frequency data in our regression models. Our results showed that
MIDAS regression using high‐frequency stock return data produced a better
forecast of GDP growth rate than the other models, and the best forecasting
performance was achieved by using weekly stock returns. The forecasting
result was further improved by performing intra‐period forecasting.
KEYWORDS
financial variable, forecast evaluation,forecasting, mixed frequencies
1|INTRODUCTION
When forecasting macroeconomic variables such as gross
domestic product (GDP) growth rate, researchers often
face a dilemma because data are not all sampled at the
same frequency. Most macroeconomic data are sampled
monthly (e.g., inflation, employment) or quarterly (e.g.,
GDP), whereas most financial variables (e.g., interest
rates and asset prices) are sampled daily or even more fre-
quently. Furthermore, asset prices are forward looking,
and they are long believed to contain useful information
about future economic developments (Stock & Watson,
2003). It is therefore interesting to examine whether or
not one can use high‐frequency financial variables to bet-
ter estimate and forecast macroeconomic variables. This
is particularly useful for the central banks, financial
firms, and any other entity whose outcome depends on
business cycle conditions and who needs to monitor the
state of the economy in real time.
However, the main challenge is how best to use such
available data—especially the high‐frequency data—in
our economic forecast, and how to effectively conduct the
time‐aggregating, such as averaging, of the high‐frequency
data to match the sampling rate of lower‐frequency data,
as time aggregationalways leads to loss of individual timing
information that might be important for forecasting.
Conventional forecasting models generally require the data
to be of the same frequency. Time aggregating, such as
averaging, of the high‐frequency data is usually practiced
to match the sampling rate of lower‐frequency data. But
time aggregation always leads to loss of individual timing
information that might be important for forecasting. Using
mixed‐frequency data in forecasting is expected to outper-
form conventional forecasting models, which generally
require data with the same (low) frequency. Hence finding
a suitable method to handle the high‐frequency data is a
crucial task for every forecaster dealing with mixed‐fre-
quency data.
One promising method currently available is the
mixed‐data sampling (MIDAS) regression model first
introduced by Ghysels, Santa‐Clara, and Valkanov
(2004), Ghysels, Sinko, and Valkanov (2007), and
Received: 18 September 2017 Revised: 3 January 2018 Accepted: 2 April 2018
DOI: 10.1002/for.2528
666 Copyright © 2018 John Wiley & Sons, Ltd. Journal of Forecasting. 2018;37:666–675.wileyonlinelibrary.com/journal/for
To continue reading
Request your trial